Skip to main content

  • Do Structural Changes Truly Precede Functional Changes in Glaucoma?

    By Lynda Seminara
    Selected By: Prem S. Subramanian, MD, PhD
    Add to My Bookmarks

    Journal Highlights

    Investigative Ophthalmology & Visual Science
    2020;61(13):5

    Download PDF

    Some evidence suggests that structural changes occur before functional decline takes place. Using structural equation models (SEMs) and biannual testing, Gardiner et al. explored whether a lag exists between true change in structure (retinal nerve fiber layer thickness [RN­FLT]) and function (standard automat­ed perimetry [SAP]) in patients with glaucoma. They found that the rate of functional change in a given interval was predictive of the rate of structur­al change in the subsequent interval, whereas the converse was not true—even though more eyes showed detect­able changes in RNFLT than in SAP.

    Data were collected for 164 patients diagnosed as having open-angle glau­coma or likelihood of developing glau­coma from the Portland Progression Project, an ongoing longitudinal study of progression and diagnostic testing in glaucoma. Rates of change were cal­culated for 318 eyes during 1,135 pairs of consecutive visits, with a mean of 207 days (range, 161-364 days) between visits. SEMs were applied to determine whether the rate of change in structure (RNFLT) or function (mean linearized total deviation [AveTDLin]) could be predicted by the concurrent or previous rate for the other modality after adjust­ment for its own rate in the preceding period.

    Analyses showed that the rate of change in AveTDLin was predicted by its own rate in the previous interval but not by the rate of RNFLT change in the concurrent or previous interval (both p > .05). Similarly, the rate of RNFLT change was not predicted by concurrent AveTDLin change after adjustment for its own previous rate. However, the rate of AveTDLin change in the previous interval was significant for predicting the current rate of RNFLT change (p = .005) and suggested a period of about six months between AveTDLin and RNFLT changes.

    Despite the lag observed in this study, the authors cautioned that the finding does not invalidate previous research suggesting that RNFLT may be detectable sooner. They encouraged the development of less variable function­al tests, along with use of alternative structural measures, to improve dam­age detection and disease prognosis.

    The original article can be found here.